¶1 delivered the opinion of the court.
¶2 A question as to the admissibility of certain evidence is certified by the Court of Civil Appeals for the First District, accompanied by a statement which shows the action to be one of trespass to try title involving in the way to be stated below the title to the tract of ten acres *108 of land marked ón the following plat, “Kinkier 10 acres according to the field notes of the deed.”
¶4 It appears from the certificate that prior to 1855 one Carl Grimmell owned the tract marked “Grimmell or Carraher 14 acres.” In that year one Huddleston conveyed to Grimmell a tract of 50 acres described as follows:
“Beginning at the Northeast corner of the 14 acres (then owned by Grimmell). Thence East 751 varas to a stake; thence 'North 293 varas; thence West 730 varas to the rim of a gully; thence down said gully to White Oak “bayou; thence down White Oak bayou to the Northwest corner of the 14 acres; thence East 408 varas to the beginning, containing 50 acres.”
¶5 It was asserted by appellees in the trial and evidence was offered to show that the call to begin at the Northeast corner of the 14 acres was a mistake for the Southeast corner; and it is stated in the certificate that, if the tract of 50 acres be thus located, it will include both tracts of 10 acres marked upon the map, while if it be located in accordance with the call in the deed it will include of those tracts only that before referred to. The bearing of this upon the controversy will appear. In 1856 Grimmell conveyed to Jacob Kinkier 10 acres described as follows:
“Beginning at the N. E. corner of Grimmell’s 14 acres. Thence East 750 varas to a stake; thence North 74% varas to a stake; thence West 750 varas to a stake; thence South 74% varas to the beginning.”
¶6 There is no dispute about the location of any line or corner of the tract of 14 acres.
¶7 Appellees, who were plaintiffs below, claim the land covered by this last description, as it is written, under bond for title from Grimmell made in 1866 and deeds in accordance therewith from his surviving widow in 1867 conveying lands including such 10 acres; and assert that there was a mistake in the older deed to Kinkier, in that it, too, called for the Northeast corner of the tract of 14 acres as the beginning when the Southeast corner was meant. Upon this contention the 10 acres sold to Kinkier would be the tract marked, “Kinkier 10 acres according to appellees’ claim.” The certificate states that the land is now in the suburbs of the city of Houston,
¶8*109 Tn support of this contention appellees, over the objection of appellants, "“that the deed referred to from Grimmell to Kinkier was the best evidence as to the location of said land and that the questions and answers thereto were immaterial and irrelevant. That the deed was the best evidence and especially as Grimmell never owned any land south of the Northeast corner of the Carraher (Grimmell) 14 acres, and his deed to Kinkier describing the land beginning at the Northeast corner of the Carraher (Grimmell) 14 acres running East and thence North and thence West and South to the beginning is plain and unambiguous, and covers land he owned at that time, and that the evidence already introduced shows that Grimmell owned the land that he described in the deed to Kinkier and did not own the land that the witness described, but that the same was owned by Mrs. McGowan,” were permitted to introduce evidence tending to show “that the 10 acres really sold, and intended to be conveyed by Grimmell to Kinkier began at the Southeast instead of the Northeast corner of the 14 acres. . . . This testimony showed that Kinkier took possession of 10 acres so described and had always claimed this as the 10 acres convened, that he sold land off of this 10 acres, and his vendees had claimed and occupied the land so conveyed to them. Much evidence was introduced along this line, all showing the intention, as understood by the parties, that the Kinkier 10 acres was located beginning at the Southeast corner of the 14-acre tract. The jury so found and the. sufficiency of the evidence to support such finding is not questioned on this appeal.”
¶9 There is no pretense that these were the pleadings or parties essential to an action for reformation of the deed. The question is: “Did the trial court err in overruling the objection of appellants, above stated, and admitting the evidence referred to?”
¶10 It is to be observed that the certificate does not state the-issues that arose in the trial of the case nor the purposes for which the evidence was offered, if they were specified, further than may be inferred from the statement which we have condensed from the certificate. It may easily be conceived that questions might arise in the trial of such an action to the decision of which the evidence stated would be relevant and admissible; but we gather from the certificate that the purpose was to show that the deed from Grimmell to Kinkier did not convey the land subsequently conveyed by Grimmell and his widow to those under whom appellees claim, but conveyed the other tract. Assuming that to have been the purpose, we answer the question in the affirmative.
¶11 There seems to be no contention that the description belongs in any class in which parol evidence would be admissible in a controversy between the parties or their privies to aid a court in solving questions left in doubt by its terms. There is no conflict in its calls to justify the disregard of descriptive particulars which appear from the language itself, when considered together, to have been mistakenly used, and the following of those which appear to be true, in order to identify the land intended. Nor does the effort to apply the description to the ground give rise to any sort of ambiguity to be removed by parol evidence showing the intention. Both on its face and in its application *110 to the ground, the description is clear and unambiguous and identifies as the land conveyed the 10 acres in dispute. It .is too well settled to admit of doubt that such a deed can not be collaterally attacked by the parties to it, or their privies, by evidence tending to show an intention different from that which its language unmistakably expresses.
¶12 But it is urged that the parol evidence pule applies only between the parties to the writing and their privies and not between strangers to it nor between one of the parties and a stranger. That this is to a large extent true of that rule as it is generally expressed, is well settled. It might be difficult by a generalization consistent with all the authorities, to define the extent to which it is true, and we shall not attempt such a task. Wigmore on Evidence, section 2446. The facts of the case stated permit the illustration of a distinction which in our opinion correct principle makes it necessary to observe. If the agreement between Grimmell and Kinkier, or their intention, as a mere fact, apart from the question as to the legal effect of the deed, were' important to any inquiry in the case, the deed would not be the exclusive evidence of such agreement or intention; but when the question is, what land did the deed convey, its legal effect between the parties is the very test invoked and it must, therefore, answer, the inquiry by its own terms, since no land was conveyed except by it and it conveys no land except that which by its terms it undertook to convey. Parol evidence, whether brought by parties or strangers, can not make it convey land which it does not purport to convey nor prevent it from conveying that which it does clearly purport to convey. Of the authorities cited by counsel for appellees, that which appears to come closest to sustaining their contention is an observation of Judge Roberts in Hughes v. Sandal, 25 Texas, 162. It is seldom that even a dictum of that eminent jurist is found to be inaccurate or careless, but if his remark should be taken as having the meaning ascribed to it by counsel, we should feel obliged to treat it only as an inaccurate dictum. The parol evidence admitted in that ease was correctly held to be admissible under the rule as to latent ambiguities, and if the further reference to the inapplicability of the parol evidence rule to contests between strangers to the deed was intended to apply t'o a description in a deed in which there was no doubt or uncertainty, it was clearly a dictum; and if it meant that strangers can, by parol evidence, prevent-such a deed from conveying land which it clearly undertakes to convey and make it convey land which it clearly does not undertake to convey, it was as clearly erroneous. Watts v. Howard, 77 Texas, 71; Powers v. Minor, 87 Texas, 88; Parley v. Deslonde, 69 Texas, 461. The description' before Judge Roberts, was one which could not be applied to the land without the aid of parol evidence properly admitted, and we feel sure he did not mean to assert such a proposition as that just stated. If it were admitted that Grimmell intended to convey the lower 10 acres as contended by appellees, that intention could not effect such conveyance nor prevent the deed, unless corrected in some proper way, from standing as the legal conveyance of the land described in it. A contrary decision would virtually repeal the statutes regulating the conveyance of lands.